Objectives
We hypothesized that ultrasonography for salivary gland stone detection would have a diagnostic accuracy similar to that confirmed by sialendoscopy, sialography, or surgery. Therefore, we evaluated the diagnostic characteristics of ultrasonography in terms of submandibular and parotid stone detection compared to confirmatory methods.
Methods
We searched PubMed, Embase, the Web of Science, SCOPUS, and the Cochrane database to October 31, 2021. The risk of bias was evaluated using the QADAS-2 tool.
Results
Ten studies involving 1393 patients were included in the analysis. The diagnostic odds ratio of ultrasonography was 162.6013 (95% confidence interval [CI] [53.9883; 489.7208] and I 2 value 81.0%). The area under the summary receiver operating characteristic curve was 0.963. The sensitivity, specificity, negative predictive value, and positive predictive value were 0.8992 (95% CI [0.8534; 0.9318]; I 2 = 79.9%), 0.9664 (95% CI [0.9290; 0.9844], I 2 = 65.6%), 0.8076 (95% CI [0.7256; 0.8694]; I 2 = 80.4%), and 0.9853 (95% CI [0.9629; 0.9943]; I 2 = 77.4%), respectively. However, high-level among-study heterogeneity (I 2 ≥ 50%) was evident, attributable to the inclusion of different glands. On subgroup analysis, significant differences in the negative predictive values (parotid gland only [0.9392], submandibular gland only [0.6718], and parotid and submandibular glands [0.8105]) were apparent. We found no significant among-study difference in the sensitivity, specificity, positive predictive value, or diagnostic odds ratio (P > .05).
Conclusion
Ultrasonography usefully detects submandibular and parotid gland stones. Ultrasonography of the parotid gland was associated with the highest diagnostic accuracy, but further clinical studies are needed.
Level of Evidence
NA Laryngoscope, 2022